Although the superior portion of the shoulder joint capsule is a clearly distinct structure, its ability to prevent vertical humeral head translation in irreparable rotator cuff tears remains controversial. The clinical results of superior capsule reconstruction (SCR) in this context are satisfactory but the radiological outcomes are poor in up to two-thirds of patients. Moreover, there is no proof that SCR is more beneficial than simply doing a subacromial debridement, a partial repair or a tenotomy of the long head of biceps. Furthermore, graft healing is not predictive of success. Given the lack of comparative studies with high level of evidence (level 1 or 2), we cannot conclude that this surgical technique (and its many variants) has a place in our treatment paradigm. Also, it seems that xenografts are no longer being used given the high rate of structural failure seen on MRI and the large number of complications. If the long head of biceps tendon is still intact at the time of surgery, it is a feasible option for SCR given the reasonable cost and minimal complications. If it is not - but the subscapularis tendon is intact or repairable - an allograft or a non-local autograft should be considered, despite its unpredictable clinical results. Surgeons should inform patients of the high rate of radiological repair failure and the worse results if a reverse shoulder arthroplasty should be needed subsequently. Level of evidence: 4.